View
0
Download
0
Category
Preview:
Citation preview
Effet du CO2 sur la circulation atmosphérique tropicaleet les changements régionaux de précipitation
Sandrine Bony (LMD, Paris)
En collaboration avec :Gilles Bellon (CNRM, Toulouse), Daniel Klocke (ECMWF, Reading),
Steve Sherwood (Univ New South Wales, Sydney),Solange Fermepin (LMD, Paris) et Sébastien Denvil (IPSL, Paris)
CMIP3 Precipitation projections
IPCC 2007
How to tackle the problem ??
What is robust ? What is not ?
Why ? What implications ?
First step : Unravel the physical
mechanisms underlying the change
in tropical precipitation pattern...
(Bony et al, in revision)
What controls the regional patternof precipitation change ?
Increased CO2 is known to affect the hydrological cycle through :
– Surface warming (e.g. Held and Soden 2006)
– Tropospheric adjustments (e.g. Gregory and Webb 2008)
How much do regional precipitation changesdepend on surface warming ?
CMIP5
A hierarchy of models, experiments, configurations (coupled ocean-atmosphere, atmosphere-only, aqua-planet..)
Model Evaluation Climate Projections
Understanding
Models :Models :ESMOAGCMAGCMAqua-planet
Experiments :Experiments :realisticidealizedpartial forcingsidealized
experiments
realisticexperiments
scenarios
Analysis Method
dynamicalcomponent
thermodynamicalcomponent
surfaceevaporation
verticaladvection
horizontaladvection
shape ofomega profile
Tropical Precipitation Projections
RCP8.5 scenario at the end 21C
+
=
Total
Thermodynamical
Dynamical
Tropical Precipitation Projections
RCP8.5 scenario vs idealized abrupt 4xCO2 expt
+
=
Total
Thermodynamical
Dynamical +
=
Total
Thermodynamical
Dynamical
RCP 8.5, end 21C Abrupt 4xCO2, ∆Τ = 4K
→ an opportunity to understand precipitation changes in RCP scenarios
dynamicalcomponent
thermodynamicalcomponent
Interpretation of regional precipitation changes
dynamicalcomponent
thermodynamicalcomponent
How would precipitation respond to global warming
in the absence of change in vertical motion ?
subsidence regimes
dP/dTs
convective regimes
16 CMIP5 models (mean and spread)
wet get wetter, dry get drier
wet get getter more robust than dry get drier
dE/dTs
Clausius-Clapeyron
0
How would precipitation respond to global warming
in the absence of change in vertical motion ?
RCP8.5(end 21C)
[mm/d] [mm/d]
DynamicalDynamicalcomponentcomponent
Evolution of regional precipitation changesin abrupt 4xCO2 experiments
T
ThermodynamicalThermodynamicalcomponentcomponent
dynamicalcomponent
thermodynamicalcomponent
How does the tropical overturning circulation
respond to increased CO2 ?
Tropical Overturning Circulation
Convective regimes Subsidence regimes
Change in circulation (%) predicted by CMIP5 models... in multiple models, experiments and configurations
1
Physical Interpretation and Time Scale of theDynamical Response to CO2 Radiative Forcing
CMIP5GCMs
1D model
Direct CO2 effect on the tropical overturning circulation :
- Robust result (CMIP5 AGCMs, 1D model (in WTG mode), ECMWF IFS operational model, SP-CAM)
- Physical interpretation
- Controlled by (ultra) fast physical processes in convective regimes (half the eq response within ~ 5 days)
ECMWF-IFS Operational Model
1xCO2
4xCO2 – 1xCO2Forecasts :
Day 5
Day 10
Day 1
Monthly-meanprecipitation
(October 2011)
Weakening of the circulation by global warming
Direct effect of CO2 forcing on large-scale vertical motions
The direct effect is : - fast - robust - not primarily mediated by sfce warming or land-sea contrasts
Explains most of the fast precipitation response to CO2 forcing
Substantially contributes to the regional pattern of long-term precipitation changes !
Implications regarding geo-engineering strategies
Next :
– use this analysis framework to interpret inter-model differences in precipitation projections (including IPSL-CM5A vs IPSL-CM5B, paleo changes, etc)
– further understand inter-model differences in the circulation sensitivity to greenhouse gases and temperature changes
– explore implications for climate variability and predictability
CONCLUSIONS
Recommended